
Replicate lets you run open-source machine learning models with a simple API. No ML infrastructure expertise required—just API calls to thousands of models.
Key Features:
Model categories:
Replicate democratizes access to ML models. You can experiment with new models within minutes instead of days of infrastructure work. For applications that need specific capabilities not available in commercial APIs, Replicate often has the solution.

The AI cloud for open-source models
Together AI provides infrastructure for running, fine-tuning, and deploying open-source models. The platform for teams that want control over their AI stack.

A unified interface for LLMs
OpenRouter provides a single API to access models from OpenAI, Anthropic, Google, Meta, Mistral, and dozens of other providers. One integration, all the models.

Build with GPT-4, DALL·E, and more
OpenAI API provides access to GPT-4, GPT-4 Turbo, DALL·E, Whisper, and embedding models. The foundation for countless AI applications.